Economics Working Paper Series Department of Economics ...

2017 OKSWP1701

Economics Working Paper Series Department of Economics

OKLAHOMA STATE UNIVERSITY http://spears.okstate.edu/ecls/

Oklahoma Oil and Gas Severance Taxes: A Comparative Analysis

Mary N. Gade Oklahoma State University Karen Maguire Oklahoma State University Francis Makamu Oklahoma State Univeristy

Department of Economics Oklahoma State University Stillwater, Oklahoma 339 BUS, Stillwater, OK 74078, Ph 405-744-5110, Fax 405-744-5180

Oklahoma Oil and Gas Severance Taxes: A Comparative Analysis 1

Abstract. Oklahoma assesses a production tax of seven percent on the extraction of oil, natural gas, and other minerals. However, since July 2002, it has taxed production from horizontal wells at only one percent for the first 48 months of production. This is a significant tax incentive relative to its neighboring states, Texas and Kansas, particularly considering the limited evidence as to the effectiveness of severance tax incentives for increasing in-state development of immobile resources. This paper empirically examines whether the severance tax incentive has encouraged horizontal development in Oklahoma relative to Texas and Kansas. Our findings indicate that the Oklahoma tax exemption has not had a significant influence on horizontal drilling. Keywords: Severance Tax, Oil and Natural Gas, Hydraulic Fracturing JEL Codes: H71, H73, Q32, Q35, Q48

Mary N. Gade Associate Professor, Department of Economics and Legal Studies, Oklahoma State University, Stillwater, Oklahoma 74078, [email protected] Karen Maguire 2 Assistant Professor, Department of Economics and Legal Studies, Oklahoma State University, Stillwater, Oklahoma 74078, [email protected] Francis Makamu PhD Candidate, Department of Economics and Legal Studies, Oklahoma State University, Stillwater, Oklahoma 74078, [email protected]

1

We would like to thank the Oklahoma State University Department of Economics and Legal Studies for providing financial support for data purchase. 2 Corresponding Author.

1

I. Introduction The development of horizontal drilling and hydraulic fracturing technologies has triggered a resurgence in oil and natural gas production in the United States. State governments levy different types of production taxes on these unconventional wells, at different rates, with a diverse mix of exemptions, deductions, and incentives. These tax policies are designed to encourage industry activity, development, job creation, and to enhance economic benefits to the state. The state of Oklahoma has carried one of the lowest effective tax rates on horizontal wells when compared to peer states. 3 The low rate is partly triggered by a four-year production tax holiday that reduces the tax on newly completed horizontal wells from seven to one percent in the first 48 months of production. This tax break has a significant impact on state revenues, but its effect on drilling activity is an open question. This paper empirically examines whether the horizontal wells severance tax incentive has encouraged horizontal drilling activity in Oklahoma relative to Texas and Kansas. Previous literature has examined the effects of the state’s political environment on oil and gas development and found that price rather than politics determined oil and gas development (Maguire 2012). On federal lands, the regulatory environment was a key factor in oil and gas leasing (Maguire 2016). 4 Other work has examined the effects of oil and gas development on regional economic outcomes such as unemployment with mixed results (Lee 2015; Munasib and Rickman 2015; Weinstein 2014; Weber, 2014; Weber 2012; Michaels, 2010).

3

Throughout the paper, references to horizontal drilling and horizontal wells, refer to those wells that use a combination of horizontal drilling and hydraulically fracturing to access unconventional oil and natural gas resources. These are commonly referred to as fracked or hydraulically fractured wells. See Fitzgerald (2012) for a detailed discussion of the economics of fracking. 4 Oklahoma, Texas, and Kansas do not have significant federal lands. (BLM, Public Land Statistics)

2

To our knowledge, our study is the first to examine the effects of Oklahoma horizontal well tax incentives. The subject is of interest as Oklahoma offers a generous exemption on horizontal production while other neighboring states, such as Texas and Kansas, do not. Providing tax incentives would be considered an effective policy for the state if it stimulates development. However, the alternative would suggest that additional natural gas production is profitable to firms regardless of the incentives and extending tax breaks only harms Oklahoma’s economy by reducing its tax revenue. The contemporary Tiebout (1956) tax competition literature devotes much attention to the question of whether or not state tax differentials have an empirical impact on the location of economic activity. Theoretically, when a jurisdiction lowers its tax rate on a mobile capital base, the net of tax return rises above that available elsewhere, and capital flows in until net of tax returns are equalized across all jurisdictions. The resulting incentive is that a lower tax rate on capital relative to surrounding jurisdictions has the potential to encourage economic activity as part of state development policy. Thus, states utilize exemptions, deductions, and other tax incentives in an effort to stimulate economic development. Most of the empirical work on state tax incentives centers on the effects of state taxation on geographically mobile capital with mixed results. Evidence ranges from a positive influence of incentives on location (Bartik 1985; Helms 1985; Bartik 1989; Papke 1991; Papke 1994; Holmes, 1998; Strauss-Kahn and Vives, 2009) to a small effect or none at all (Schmenner 1982; Plaut and Pluta 1983; Carlton 1983; Schmenner, Huber and Cook 1987; Blair and Premus 1987; Dabney 1991; Tannenwald 1996; Lee 2008). Literature reviews also suggest that the results are somewhat ambiguous (Wasylenko 1999; Buss 2001; and Arauzo-Carod, et al 2010).

3

Examination of the effects of state severance tax incentives on oil and gas drilling and production activity requires an alternative perspective. Nonrenewable natural resources are geographically immobile. Firms engaged in oil and gas production cannot change location to the extent that their main capital consists of the immobile reserve base. Therefore, state severance tax preferences cannot encourage relocation across jurisdictions. Firms can respond, however, by altering the level and timing of extraction as state tax policy changes. The introduction of horizontal drilling along with hydraulic fracturing technology has expanded the ability of firms to profitably recover natural gas and oil from unconventional sources. 5 However, horizontal drilling comes with higher costs relative to conventional drilling. A horizontal well can cost between 25 to 300 percent more to drill and complete. Due to the difference in cost structure, horizontal drilling is restricted to geological plays with low permeability where vertical wells would not be considered economically viable (Helms, 2008). There have been few empirical studies that look at how firms in nonrenewable resource industries respond to state tax incentives. They have included simulation and/or econometric studies of the relationship between a specific tax policy and natural resource exploration and production. For example, Kunce, et al (2003) simulate the effects of state tax policy changes in Wyoming on both the level and timing of exploration and output in the state. The simulation is based on the standard model of natural resource supply from Pindyck (1978) and estimates exploration costs, production of reserve additions, and extraction costs for Wyoming while incorporating the specific tax parameters that are of interest to the industry. In contrast, Leighty and Lin (2012) estimate field-specific cost functions based on cost and production data from

5

INTEK, Inc., 2011, Review of emerging resources: U.S. shale gas and shale oil plays, Energy Information Administration, DOE. Web accessed 05 August 2015. http://www.eia.gov/analysis/studies/usshalegas/pdf/usshaleplays.pdf

4

Alaska’s North Slope and use those estimated equations to model production decisions and simulate the impact of tax policy on the production rate. The conclusions drawn from these studies and others suggest that the volume of oil and natural gas production and drilling activity is insensitive to severance tax rate changes (see also Helmski-Oskou, et al 1992, Kunce 2003, Chakravorty, et al 2010, Kaiser 2012). In this paper, we build on this existing literature by examining whether Oklahoma’s severance tax reduction has led to disparate effects for horizontal versus conventional drilling as compared with its neighboring states. These states, particularly Texas, share a similar oil and gas resource potential, which we will use in this paper in order to isolate the effect of the tax policy on development from differences in development due to resource disparities. Our findings suggest that the Oklahoma tax exemption has not increased horizontal drilling activity.

II. Background 11.1 Oil and Natural Gas Resources Oklahoma, Texas and Kansas are not only among the states to apply the lowest oil and gas production taxes in the nation, but are also among the major contributors of oil and gas in the country. These states are selected in this study for their regional proximity as well as their long history of oil and gas production. For the period 1982 to 2013, Texas had the highest production (onshore) of natural gas, Oklahoma comes in third position and Kansas is seventh. 6 The Hugoton gas field which contains one of the largest producing natural gas fields in the U.S. is located in southwestern Kansas and includes parts of the Texas and Oklahoma panhandles. In 2013, Texas accounted for 29 percent of the country’s marketed gas production, well above any other state.

6

U.S. Energy Information Administration. 2015. Natural gas gross withdrawals and production. Authors’ calculations. http://www.eia.gov/dnav/ng/ng_prod_sum_a_epg0_fpd_mmcf_a.htm

5

Oklahoma accounted for 7.1 percent of the U.S. natural gas production and 8.4 percent of the U.S. marketed production in 2013. Oklahoma is one of the leading states in terms of the number of gas wells drilled, only behind Ohio, Pennsylvania, West Virginia and Texas. From 1995 to 2013, Kansas ranked ninth in the number of producing gas wells. Texas is also the leading oil producing state with production levels sometimes exceeding the federal offshore areas. In 2013, Kansas ranked tenth in crude oil production among the 50 states, while Oklahoma ranked fifth in crude oil production in the nation in 2014. 7 II.2 Horizontal drilling and Hydraulic fracturing Horizontal drilling combined with hydraulic fracturing technology allows producers to develop deposits of oil and natural gas that are trapped in deep shale and tight sands formations often one mile below the surface. To access the resources, producers drill vertically until they reach the reservoir, the kickoff point, and then the wellbore starts curving horizontally. The drill pipe is removed and replaced by a steel casing pipe through which cement is pumped in order to isolate the wellbore from any freshwater source. Hydraulic fracturing consists of pumping a mixture of water, sand and chemicals under controlled conditions into deep underground shale or tight sands formations. The injected sand and fluid remain in the rock to leave cracks open so that when the pump pressure is released the oil or natural gas can flow into the horizontal casing and then up to the wellbore. 8 The concept of hydraulic fracturing was developed in 1891, and implemented for the first time in Texas around 1929, but it only become commercially viable in the 1980s. The cost of horizontal drilling is generally higher than the conventional vertical drilling, up to 300 percent.

7

http://www.eia.gov/state/ American Petroleum Institute. 2009. Hydraulic fracturing operations, well construction and integrity guidelines. API Guidance document HF1. Web accessed 09 September 2015. http://www.shalegas.energy.gov/resources/HF1.pdf

8

6

However, when financially viable a horizontal drill can produce 2.5 to 7 times the rate and reserves of conventional wells and allow producers to access once economically infeasible resources (King, 1993). Since the early 2000s the price of natural gas and other fuels have been steadily rising making more expensive and complicated methods like hydraulic fracturing more attractive. 9 It should be noted that gas prices have declined since 2008 and oil prices have fallen significantly recently, leading to a reduction in oil and gas development activity, but this decline occurred largely after our sample period, which ends in 2013. (Arezki and Blanchard, 2014; Asche, Oglend and Osmundsen, 2012; Baffes et al., 2015).

II.3 State Severance Tax Structures: Oklahoma, Texas, Kansas Tax structures are not easy to compare as they vary across states and at different levels of government authority, i.e. federal, state, and local jurisdictions. To address this problem, the literature often makes use of the effective tax rates, taxes divided by production value, in order to account for various form of tax incentives granted by states. For example, including severance, property, income and sale taxes, Oklahoma, Texas and Kansas have in FY 2010-2011 effective tax rates between 7.4 and 8.4 percent. 10 Although their effective tax rates are similar, there are major differences in their tax structures. For example, Kansas levies high property taxes while Oklahoma does not, and Texas does not levy a corporate income tax while Oklahoma does. In addition, counties can impose property taxes on drilling companies, which are independent of the well economic value, but the revenue from taxation on structures and equipment are generally not as large as the revenue obtained from the severance tax (Kunce & Morgan, 2005). Also, state and federal corporate income taxes are levied on the firm profit and have very little effect on the

9

U.S. Energy Information Administration. 2015. U.S Natural Gas Prices. Wellhead price from 1922 to 2012. Carey, M. May 11, 2014. Effective severance tax rates [Memorandum]. Colorado Legislative Council. Denver, CO.

10

7

firm’s decision to explore and produce. They are more likely to affect future drilling investments and output, rather than the actual number of drilling permits issued (Deacon, 1993). States grant numerous exemptions and credits to oil and gas producers (Kunce, 2003). For example, Oklahoma, Texas and Kansas offer exemptions for inactive wells for at least a certain period after the date of certification of two or three years of inactivity. Their tax structures also differ on other features. The Oklahoma gross severance tax rate on gas is 7 percent when the price of gas exceeds $2.10 per thousand cubic feet (Mcf). The state offers exemptions on severance tax for recovery projects, inactive wells (reestablished production), horizontally drilled wells, new discovery wells, and seismic exploration. Since 2002, Oklahoma has exempted gas production from horizontal drilling to claim only 1 percent for the first four years of production. 11 It is the most generous exemption provided by the state which can be claimed by drilling companies for unlimited production for the first forty-eight months regardless of the current market price. Like Oklahoma, Texas has taxed oil and gas production at a rate of 7.5 percent of the market value of gas produced and 4.6 percent of the market value of oil produced. 12 (See Appendix: Table 2A for detailed information on state tax incentives for Kansas, Oklahoma, and Texas.) The state has provided natural gas tax incentives since 1989 in the form of an exemption for production of gas with higher drilling costs (“high cost gas”) and inactive wells, both for a period of ten years. The tax benefits have been extended several times and made permanent in 2003. More important to our identification, the tax code does not include production tax exemptions for horizontally drilled wells in oil and gas production. 13

11

Oklahoma Statutes. Chapter 68 §1001 (E). Texas Tax Code. Chapter 201 section 52 (A). 13 Other exemptions comprise gas production from low producing wells on and after 2005 and gas production in association with geothermal energy production after 2009. 12

8

Kansas has a severance tax rate of 8 percent on the gross value of oil and gas, implemented on April 1983. 14 The state provides two years of severance tax exemption for oil and gas extracted from new pool and fifteen years tax exemption for inactive wells. Gas production for enhancement recovery projects, new discovery wells, and seismic exploration are exempt for seven years but subject to the market price of gas. Like Texas, Kansas does not provide any specific tax incentives for horizontal drilling. Overall, and essential to our identification strategy, unlike Oklahoma, Texas and Kansas do not provide tax incentives that specifically target horizontal drilling. Also, these neighboring states have not enacted a major change in their tax code for other types of production. Texas has not revised chapter 201, its administrative code on tax exemption for natural gas, since 1997, except for minor modifications in 2005 and 2009 (House Bill 2161 and Senate Bill 997). Similarly, Kansas has not introduced new exemptions since 1998 (Kansas statutes 79-4217). There are myriad of minor regulatory changes, but we didn’t find any major regulatory changes that would confound our findings. 15 While it is clear that increases in gas production lead to increases in tax revenue, it is not clear that the reduction in Oklahoma’s severance tax rate for horizontal wells will lead to additional development. This paper focuses on the empirical question of whether the tax credit on horizontal drilling in Oklahoma led to increased drilling permits relative to Texas and Kansas.

III. Data

14

Kansas Statutes. Chapter 79 Article 42 § 17(B). Oklahoma Corporation Commission: Chapter 10 - Oil and Gas conservation (OAC 165:10-3-10) in http://www.occeweb.com/rules/rulestxt.htm. Railroad Commission of Texas: Chapter 3 – Oil and Gas Division (TAC 16:1.3) in http://www.rrc.state.tx.us/legal/rules/current-rules/; Kansas Corporation Commission – General rules and regulations for the conservation of crude oil and natural gas (KAR 82:3) in http://kcc.state.ks.us/conservation/cons_rr_091615.pdf 15

9

For our analysis, the number of oil and natural gas permits issued serves as our measure of oil and gas development. The steps in the development process include permitting, drilling, completion, and production. In each step after the initial drilling permit is issued, firms face additional costs and resource constraints in order to complete the well development process. For this reason, the number of permits issued provides the best measure of the influence of a tax policy on a firm’s initial oil and gas development decision without the confounding effects of subsequent changes in development costs and other constraints facing these firms. The permit provides the firm with authorization to drill in the designated location abiding by any state restrictions regarding drilling methods. 16 Therefore, the measure provides a complete set of oil and gas development activity; for each oil or natural gas well drilled in a state, the state’s oil and gas commission requires a permit. The drilling permit data for the analysis were collected from the relevant state oil and gas and or geologic agency (The Oil and Gas Division of The Oklahoma Corporation Commission, The Railroad Commission of Texas, and The Kansas Geologic Survey). 17 Each drilling permit is categorized by the relevant agency as conventional or horizontal. In addition, permits are designated as oil, natural gas, or simply oil and gas. 18 Data are collected for the sample period 1995-2013 and aggregated to the month-county level for each county during the sample period. In addition to permit data, we also collected data on monthly oil and natural gas prices by state

16

The permits typically require that drilling begin within 6 months to a year from the issue date and expire if not used. The number of permits issued and the number applied for are essentially the same. For example in Oklahoma over the sample period only 20 permits were rejected. In order to eliminate those few cases where permits were denied, we used only approved permits in the measure. 17 Permits for oil and gas drilling and recompletion were considered. Permits for other wells such as injection wells are excluded from the analysis. For Texas, only onshore permits are analyzed. 18 For permits for oil and gas drilling that were not categorized separately as oil or natural gas, we defaulted to a joint oil and gas category.

10

from the Energy Information Association (EIA). 19 Data on population and personal income at the county-year level were collected from the Bureau of Economic Analysis (BEA). Lastly, information on whether a county is a metro or non-metro and whether the counties were consistent oil and gas producing counties were collected from the Economic Research Service of the United States Department of Agriculture (ERS-USDA).

IV. Empirical Specification The paper empirically examines whether the severance tax reduction in Oklahoma in 2002 has led to an increase in the amount of horizontal drilling relative to Texas and Kansas. The empirical analysis is focused on measuring the influence of the horizontal drilling severance tax policy in Oklahoma on the number of drilling permits issued and how this may have differentially impacted horizontal and conventional drilling. The main specification of the fixed effects model is: 𝑌𝑌𝑖𝑖𝑖𝑖 = ∝ +𝛽𝛽1 𝑅𝑅𝑖𝑖𝑖𝑖 + 𝛽𝛽2 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑗𝑗𝑗𝑗 + 𝛽𝛽3 𝐼𝐼𝐼𝐼𝐼𝐼𝑖𝑖𝑖𝑖 + 𝛽𝛽4 𝑃𝑃𝑃𝑃𝑃𝑃𝑖𝑖𝑖𝑖 + 𝛽𝛽6 𝑖𝑖 + 𝛽𝛽7 𝑡𝑡 + 𝜀𝜀𝑖𝑖𝑖𝑖 Where i = county, j = state, t = month, and y = year (Y) represents the monthly county number of permits issued. (𝑅𝑅) is an indicator of the

Oklahoma tax policy of interest. (Price) is a state-month real oil or natural gas price. (Inc) is county-year real personal income. Lastly, (Pop) is the county-year population.

19

For oil prices, we used the EIA first purchase price, while for natural gas prices we used the EIA citygate price; first purchase or wellhead prices for natural gas were discontinued in 2013.

11

Figure 1: Share of Permits (Horizontal/Conventional) by State

The difference-in-difference (DID) technique that is used relies on the assumption that changes in drilling permits over time would have been the same in both the treatment, Oklahoma, and control states in the absence of the intervention. Figure 1, above, demonstrates that from 1995 through the early 2000s, the share of conventional and horizontal drilling permits approved was constant for all three states. This is due largely to the absence of hydraulic fracturing technology prior to the mid-2000s and its rapid adoption throughout the United States thereafter. The extent to which the rise in horizontal drilling in Oklahoma demonstrated in Figure 1 is due to this technological advancement and the degree to which it was influenced by the tax policy in Oklahoma is therefore an empirical question. By differencing out the availability of hydraulic fracturing technology, which was also available in Kansas and Texas we are focused

12

on identifying the Oklahoma tax policy impacts. Specifically, the county and month fixed effects specification controls for all time invariant county characteristics, and the effects of seasonal and annual national economic and policy changes that influenced both Oklahoma and the control states in the sample. Still, the possibility of endogeneity exists due to unmeasured county-level heterogeneity in resource potential, which may lead to changes in the number of permits issued and the implementation of the tax policy. In order to control for resource availability, we have used an alternative sample to focus on geologically similar regions across states as designated by the United States Geologic Survey (USGS). In addition, to address concerns over unmeasured local economic factors, we have controlled for dynamic changes in personal income and population. Results for the full sample and the alternate specifications are provided below. V. Results The findings in Table 1A indicate that there is no consistent effect of the Oklahoma tax policy for horizontal drilling on the number of horizontal permits issued after the implementation of the policy in 2002. Despite the dramatic increase in horizontal drilling demonstrated in Figure 2, the results suggest that the tax break was not a determining factor, rather it was oil prices that had a significant influence. The rise in horizontal drilling was not limited to Oklahoma. Figure 2 also demonstrates marked growth in the number of horizontal drilling permits in Texas, which did not have the policy. Clearly, the technological expansion of horizontal drilling crossed state lines and led to growth in permitting in both Texas and Oklahoma.

13

Figure 2: Horizontal and Conventional Permits by State

Unexpectedly, the findings indicate that the number of conventional drilling permits dropped in Oklahoma as compared to its neighbors after the implementation of the tax policy. This drop is demonstrated in Figure 2. Each of the states experiences a sharp decline in conventional permits in 2008, presumably as the result of the Great Recession. However, for both Texas and Kansas, conventional drilling activity recovers and resumes an upward trajectory through 2011, Oklahoma does not. The findings in Table 1B, Column 1 indicate that on average the tax policy led to a decline of approximately one permit per county month. This is an economically significant finding, because the mean county-month permits issued is 3.5. (See Appendix Table 1A for Summary Statistics). The findings for both conventional and horizontal permits are robust

14

to several modifications of the sample and alternative scaling of the dependent variable, presented below. 20

20

In addition, natural gas prices were also considered and the statistical and economic significance of the policy intervention were consistent when the analyses were conducted using natural gas rather than oil prices. Results available upon request.

15

Table 1: Main Specification Dependent Variable: # of MonthCounty Oil/Gas Permits Policy Intervention Oil Price (State/Month) Personal Income (Cty/Year) Population (Cty/Year)

A: Horizontal Drilling Permits All OverShare Counties lapping (3) Regions (1) (2)

Exclude Oil Permits (4)

Oil And Gas Counties (5)

B: Conventional Drilling Permits All OverShare Counties lapping (3) Regions (1) (2)

Exclude Oil Permits (4)

Oil And Gas Counties (5)

0.0781 (0.364)

-0.132 (-0.361)

0.0626*** (3.542)

0.168 (0.829)

0.0355 (0.153)

-1.010*** (-3.713)

-0.685* (-1.807)

-0.0869*** (-3.444)

-0.577* (-1.927)

-1.153*** (-3.837)

0.0055*** (4.884)

0.00653*** (2.684)

0.00032*** (4.011)

0.0043*** (3.984)

0.0062*** (4.938)

0.0200*** (7.275)

0.0191*** (5.878)

0.0005*** (3.526)

0.0177*** (8.010)

0.0222*** (7.278)

-3.04x10-07 (-1.495)

-5.00x10-07 (-1.473)

-3.94x10-09 (-0.648)

-3.12x10-07 (-1.543)

-3.84x10-07 (-1.574)

1.94x10-07 (1.248)

8.42x10-08 (0.811)

1.62x10-09 (0.293)

-1.18x10-07 (-0.479)

2.45x10-07 (1.251)

3.37x10-05 (1.356) 95,076 0.055

8.05x10-05* (1.805) 36,480 0.132

6.03x10-07 (0.918) 95,076 0.067

3.48x10-05 (1.402) 95,076 0.052

4.12x10-05 (1.414) 85,044 0.063

-1.9x10-05* (-1.729) 95,076 0.028

-1.28x10-05 (-1.329) 36,480 0.043

-4.11x10-07 (-0.775) 95,076 0.021

1.56x10-05 (0.607) 95,076 0.043

-2.4x10-05* (-1.697) 85,044 0.031

Observations R-Squared Number of 417 160 417 417 373 417 160 417 417 373 Counties Note: Robust t-statistics in parentheses. *** p